Data-driven COVID-19 growth prediction
2nd International Conference on Computing and Data Science, CDS 2021
; : 64-72, 2021.
Article
in English
| Scopus | ID: covidwho-1364914
ABSTRACT
COVID-19 disease become the most influential public health event in 2020. It has affected more than two hundred countries and regions. The prediction and analysis of the epidemic is extremely important. It can help governments and international organizations control the development of the epidemic. In our study, we use three different models, namely, autoregressive integrated moving average (ARIMA), multilayer perceptron (MLP) and long short-term memory (LSTM). We made predictions based on the data from 20 countries, which reported the largest confirmed cases until June 30, 2020. We find that the ARIMA model achieves the lowest error. For Turkey, the forecast root mean square error (RMSE) from June 3 to June 30, 2020 is only 95.049. Our research will help the government to follow up the future development of COVID-19 and make decisions. It will also play a role in the outbreak of major diseases that may appear in the future. © 2021 IEEE.
Full text:
Available
Collection:
Databases of international organizations
Database:
Scopus
Type of study:
Prognostic study
Language:
English
Journal:
2nd International Conference on Computing and Data Science, CDS 2021
Year:
2021
Document Type:
Article
Similar
MEDLINE
...
LILACS
LIS